Skip to main content

Full text of "Archean crust-mantle geochemical differentiation"

See other formats


Geological Sciences, University of California, Santa Barbara, CA 93106 

Isotope measurements on carbonatite complexes and komatiites can provide 
information on the geochemical character and geochemical evolution of the 
mantle, including sub-continental mantle. Measurements on young samples es- 
tablish the validity of the method. These are based on Sr, Nd and Pb data 
from the Tertiary-Mesozoic Gorgona komatiite (1, 2 ) and Sr and Pb data from 
the Cretaceous Oka carbonatite complex (3, 4). In both cases the data describe 
a LIL element-depleted source similar to that observed presently in MORB. 

Carbonatite data have been used to study the mantle beneath the Superior 
Province of the Canadian Shield one billion years (lAE) ago. The framework 
for this investigation was established by Bell et al . (3) who showed that 
large areas of the province appear to be underlain by LIL element-depleted 
mantle (87Sr/86Sr = 0.7028) at 1 AE ago. Additionally Bell et al . found four 
complexes to have higher initial Sr ratios (87Sr/86Sr = 0.7038), which they 
correlated with less depleted (bulk earth?) mantle sources, or possibly crus- 
tal contamination. 

We have determined Pb isotope relationships in four of the complexes 
studied by Bell et al . In favorable cases the carbonates from the complexes 
yield negligible jji situ radiogenic Pb corrections {\x = 0.01 - 2), allowing 
accurate determination of initial ratios. Initial ratios for six samples 
from three of the complexes with 0.7028 initial Sr ratios (Firesand, Prairie 
Lake, Killala) plot along a regression line given by 207pb/204pb = 0.128 
206p5/2oupb + 13.186, with 206pb/20upb varying from 16.48 to 17.08. The data 
plot distinctly below crustal Pb evolution curves as given, for example, by 
Stacey and Kramers (5). The slope of the regression line, 0.128, differs 
significantly from the value expected from contamination with 1.0 AE Pb 
(0.0725) or 2.7 AE Pb (0.185). The carbonatite regression line plots to the 
left of the modern MORB regression line and has a slightly greater slope, 
apparently describing Pb isotope relationships for a billion-year-old MORB- 
like mantle source. The Pb and Sr data agree in suggesting that LIL element- 
depleted mantle existed beneath large areas of the Superior Province one 
billion years ago as far inland as the present-day Lake Superior region. Pb 
data from a fourth complex (Lake Nemegosenda, ^"^Sr/^^Sr = 0.7038) plot above 
the crustal Pb evolution curve, and agree with the Sr data in indicating 
either crustal contamination, or origin inmore LIL-enriched mantle. 

In contrast to the above results, isotopic data from 2.7 AE rocks of the 
Superior Province suggest that depleted mantle was of limited extent, and not 
sufficiently aged to have acquired an isotopic signature at that time. Sr 
data from the alkaline complex at Poohbah Lake (3) plotted nearly in the 
"bulk earth" field in a Sr evolution diagram (^^Sr/^^Sr = 0.7012) rather than 
below the field as in the case of the Oka and most of the billion year old 
complexes. Zindler et al . (6) showed that initial I'+SNd/i^'+Nd ratios in 2.7 
AE komatiites and tholeiites of Munro Township, Albitibi District plot on 
the chondrite evolution curve at 2.7 AE, rather than above the curve as is 
observed for rocks from depleted mantle, e.g., MORB. 

More recently, a Pb isotope study has been started on rocks from Munro 
Township (2). Sampling to date includes komatiite, tholeiite and sulfide 
ores. The komatiites are taken to represent mantle isotope relations, while 
the ores should characterize crustal isotopic compositions. When the data 
are plotted in a 207pb/204pb - 206Pb/20upb diagram (Fig. 1) they closely fit 
an isochron giving an age of 2.65 AE. Statistical analysis of the regression 
line yields a Y axis intercept of 12.101 ± 0.035, a deviation within the 2a 



Tilton 6.R. 

errors of the individual ^oypb/aoitpb ratio measurements. This agreement is 
taken to indicate that all members of the suite have nearly identical age and 
initial Pb isotope ratios. The Pb data thus appear to be consistent with the 
SR data from Poohbah Lake and Nd data from Munro Township in failing to iden- 
tify a LIL depleted mantle source for the komatiites and tholeiites 2.7 AE 
ago. An analogous case for Pb isotopic data from the Fennoscandian Shield of 
Finland was given by Vidal et al . (7), with the exception that the Finland 
regression line is systematically displaced above the Munro Township line, 
as shown in Fig. 1. The Finland suite includes granitic rocks that plot 
along the isochron with the komatiites and tholeiites. The only granitic Pb 
isotope data from the Abitibi District available so far are given in an 
abstract by Gariepy et al. (8), who report "large" variations in 207p5/20'tpb 
ratios in unmetamorphosed plutons. The contrast between the granite and ore 
data may indicate that the ores average out differences between individual 
plutons. Further isotopic studies of Pb are underway in the granitic rocks 
from Munro Township. 

Although evidence for depleted mantle has been observed in Nd data in 
3.5-3.8 AE rocks in other shield areas, Nd, Sr and Pb data suggest that de- 
pleted mangle originated ca. 2.7-3.0 AE ago in several areas beneath the 
Superior Province in the Canadian Shield. The relative abundance of depleted 
mantle on a world-wide basis in Archaean time remains to be definitively 
answered in future work. 


(1) Echeverria L.M. (1980), Tertiary or Mesozoic komatiites from Gorgona 

Island, Colombia: Field relations and geochemistry. Contrib. Mineral . 
Petrol. 73 , p. 253-266. 

(2) Tilton, G.R., (1983), Evolution of depleted mantle: The lead perspective. 

Geochim. Cosmochim. Acta 47 , In press. 

(3) Bell, K., Blenkinsop, J., Cole, T.J.S. and Menagh, D.P. (1982), Evidence 

from Sr isotopes for long-lived heterogeneities in the upper mantle. 
Nature 298 . p. 251-253. 

(4) Granenfelder, M.H., Tilton, G.R., Bell, K. and Blenkinsop, J. (1982), 

Lead isotope relationships in the Oka carbonatite complex, Quebec. 
EOS 63 , p. 1134 (abstract). 

(5) Stacey, J.S. and Kramers, J.D. (1975), Approximation of terrestrial lead 

isotope evolution by a two-stage model. Earth Planet. Sci. Lett. 26 , 
p. 207-221. 

(6) Zindler, A., Brooks, C, Arndt, N.T. and Hart S.R. (1978), Nd and Sr iso- 

tope data from komatiitic and tholeiitic rocks of Munro Township, 
Ontario. U.S. Geo!. Surv. Open-File Report 78-701 . p. 469-471. 

(7) Vidal. P.H.. Blais. S.. Jahn, B.M., Capdevial, R. and Tilton. G.R. (1980), 

U-Pb and Rb-Sr systematics of the Suomussalmi Archaean greenstone belt 
(Eastern Finland). Geochim. Cosmochim. Acta 44 . p. 2033-2044. 

(8) Gariepy. C, Dupre, B. and Allegre, C.J. (1982), Lead isotopic composi- 

tion in K- feldspars from the Abitibi greenstone belt and the genesis of 
the Archaean crust. EOS 63 , p. 367 (abstract). 

16.5 - 

16.0 - 


O 15.5 


£ '5.0 




• Galena 
® Other Sulfides 
® Komatiites 
o Tholeiites 


13.0 14.0 15.0 16.0 

17.0 18 19.0 20.0 21.0 


— I n: 
-1. rn 

-' S> 

r+ 2 


cr> cz 

XI -H 

22.0 23.0 24.0 




I— I 



Figure 1. Pb isochron diagram for eastern Ontario samples. The dashed 
line IS the regression line described by the Finnish data of Vidal et al . (7)